AI Safety Reading Group - Reasoning and Learning Lab

Winter 2018 Upcoming meetings

Reading List

When	Where	To be discussed	Led By
Jan 11, 2:30-4pm	Staff Lounge	The Off Switch Game [Paper], [SummarySlides]	Khimya
Jan 18, 3:00-4pm	MC 437	Concrete Problems in AI Safety [Paper]	Arushi
Jan 25, 3-4 pm	MC 437	Inverse Reward Design [Paper]	Michael
Feb 15, 3-4 pm	MC 437	Robust Adversarial Reinforcement Learning [Paper]	Jessie
Feb 22, 3-4 pm	MC 437	Safe/High Confidence Policy Improvement in a Batch Mode Setting	Paul
Mar 1, 3-4 pm	MC 437	Reinforcement Learning with a Corrupted Reward Channel [Paper]	Khimya
Mar 8, NA	NA	Cancelled Due to Study Break
Mar 15, NA	NA	Cancelled Due to RL MidTerm
Mar 22, 3-4 pm	MC 437	Algorithms for CVaR Optimization in MDPs [Paper]	Rez
Mar 29, 3-4 pm	MC 437	A Distributive Perspective on RL [Paper]	Riashat
Apr 5, 3-4 pm	MC 437	Adversarial Risk and the Dangers of Evaluating Against Weak Attacks [Paper]	Michael
Apr 12, 3-4 pm	MC 437	Reading Exercise: AI Alignment [Paper]	Cancelled
Apr 19, 3-4 pm	MC 437	Shared Autonomy via Deep Reinforcement Learning [Paper]	Khimya

When: Weekly meetings on Tuesday, Where: Grad Lounge
Papers Discussed:
- A Comprehensive Survey on Safe Reinforcement Learning [Paper]
- Should Robots be Obedient? [Paper]
- Enabling Robots to Communicate their Objectives [Paper]
- Inverse Reward Design [Paper]